PromptPwnd: Prompt Injection Vulnerabilities in GitHub Actions Using AI Agents - Aikido
Scope of the Vulnerability Pattern. Workflows are at risk if they: Use AI agents including: Gemini CLI; Claude Code Actions; OpenAI Codex ...
Read the original at Aikido Security ↗
- Topics
- CybersecurityAI Agents & Automation
- Mentions
- AnthropicClaudeGeminiGoogleOpenAI
- Tracked via
- “AI Risks”